The overall goal of the Neuroimaging core quantitative analysis of brain structures, measured by MRIs acquired at two sites in Southern acquired at two sites in Southern and Northern California, using identical protocol which permit combined use of data.. This data will be used by all of the other Cores and Projects. During the previous funding period., the Core developed a number of image segmentation, voluming, and localizing tools, all of which have been incorporated into a semi-automated user friendly, software package. The major funding of this Core has been reduced cortical gray matter and hippocampal volume in subjects with lacunes. Several of these subjects have been shown at autopsy to have ischemic change, but no evidence of Alzheimer's disease. Taken together with the PET results of reduced cortical glucose metabolism (Project 1), and MRSI results of reduced cortical [NAA] (Project 2), these MRI results emphasize cortical changes in subjects with lacunes. However, it is important to replicate these findings in a larger sample size, follow the subjects to autopsy, and co-register MRI results with pathology. However, it is important to replicate these findings in a larger sample size, follow the subjects to autopsy, and co-register MRI results with pathology. Other findings include: differences between D and D+L subjects in: the pattern of gray matter loss (measured using automated Talairach transformation and region selection), and effects of APOe4. Further development of processing software, will emphasize automation; validation will be performed using user-identified structures and autopsy confirmation (Project 3). We will use an IntellX computer and software (with J Haller) for automated voluming of hippocampus and other structures. Furthermore, we will develop automated software (with P Thompsen and A Toga) for voluming and segmentation. The following analysis will be performed: 1) Neuroradiological identification of lacune location. 2) Tissue segmentation: Brain MRI pixels will be classified into the following tissue types: Cortical gray matter,, subcortical gray matter, white matter, white matter lesions WML, lacunes, ventricular and sulcal cerebrospinal fluid. 3) Hippocampal voluming. 4) Transformations of MRI data space to identify the location of segmentation data including lacunes and white matter lesions. All data will be conveniently available to investigators of the PPG for correlation with clinical, neuropsychological, PET, MRSI, and autopsy data.